Urban Expansion and Vegetation Cover Change in and Around Jimma Town Since 1990

Urbanization is the renovation of rural society into an urban society as a result of socio-economic and political growth leading to foundation and expansion of urban agglomerations along with changing land use patterns. The main aims of this study was Investigating the extent of vegetation loss through urban expansion in and around Jimma Town over a 30 period of 6 years interval by using Geographic Information System (GIS) and Remote Sensing (RS) techniques. To achieve this objectives, Data was obtained from earth explorer (USGS) 6 years interval from1990 to 2020 of the study area. Downloaded image were extracted and each layer stacked together using the digital image-processing software ENVI 5.2. The Processed images were classified using supervised classification Algorithms into 5 hierarchical classes; Built-up area, Vegetation, Agricultural land, Grass land, and wetlands based on a modified classification scheme. Change analysis was also undertaken by applying post-classification change detection procedures. Accuracy of the image classification was assessed using error matrix, overall accuracy and kappa coefficient. The change analysis result revealed that the LULC have shown both positive and negative significant changes. Built-up were the top LULC that experienced positive change; whereas grass land, vegetation, agricultural land and wetlands have substantially declined. An important implication of the observed changes is that rapid urban expansion, compounded by poor urban planning is leading to enormous losses of key ecosystems such as wetlands and natural vegetation. The consequence of this rapid ecological degradation could potentially impact ecological functioning and environmental sustainability in and around Jimma city. Therefore, critical system thinking is required to address these complex problems in the study area and areas of rapid urbanization elsewhere in the country. of people from neighboring city and unable to compete the land lease price. These results are collaborate with (te Lintelo et al. Mosammam et al. 2017) who reported that the rapid urban population is a key challenge of the twenty-first century. From land cover classes namely built up shows an increasing trend throughout the study periods (Fig. built-up area shows increasing trend from 1990 to 2002 covering an area of 66.897 km 2 (13.16%) in the year 1990 and 129.2985 km 2 (25.45%) in the year 2002 and eight years later this land cover class increased to 385.83 km 2 (75.95542%) in the year 2020. The trends of agricultural land, grassland, and vegetation were decreased throughout the study periods except for the former class in 2002 but the rate of decrease was different. The former class expanded to forest area following rapid urban expansion near to town center on croplands in 2002. However, a built-up area rapidly replaced agricultural land following massive constructions, This finding is in line with Asmat and Zamzami(2012) and Esayas (2013) which explain about increased built-up area due to rapid, unstructured and unplanned development. Extreme decline of grassland and agricultural land was occurred in the year 2002 to 2020. This declining trend of the land is due to increasing land requirements for house construction, which is arising from rapid population growth and uncontrolled response by the government (Lejano and Bianco 2018), which enhances the problem of informal house constructions. These findings are supported by Roberts and Okanya (2018), who reported urban informality, is the outcome of either housing deficit or unaffordable cost of formal house rent. It is observed that the vegetation cover has increased from 40.068 km 2 (7.88%) in the year 2002 to 47.9002 km 2 (9.4294%) in the year 2015. On the other hand, the wetland of the study area shows a decreasing trend from 1990 to 2020 covering a total area of 69.518 km 2


Introduction 1.Back ground
Urbanization is the spatial concentration of people and economic activity. It can also be defined as the transformation of rural society into an urban society as a result of socio-economic and political growth leading to formation and expansion of urban agglomerations and city centres along with changing land use patterns. As cities are growing, the demand for resources like water, land etc. has grown proportionally to the growing rate of the urban population. This causes cities to be experiencing increasing signs of environmental stress, notably in the form of poor air quality, excessive noise and traffic congestion (Tirziu, 2020). The loss of this natural vegetation has great implications such as destruction of wildlife habitat, depreciation or outright wiping off of genetic pools, loss of food and medicinal herbs, promotion of desertification and drought among others and the building up of greenhouse gasses. Satellite image data provides the potential to obtain land cover information from inaccessible locations at more frequent intervals and also more economical than those obtained by traditional methods (Martin & Howarth, 1989;Trotter, 1991;Treitz, Howarth & Gong, 1992). The advantages of satellite imagery as compared to aerial photography include regular repeat coverage, recovering data from the same area at the same time of the day, consistence scale and look-angle, and lower cost (Kressler & Steinnocher, 1996).Planners and designers therefore need efficient tools to quantitatively evaluate and compare the impact of alternative plans and designs so that more informed development Vegetation cover plays an important role in the wellbeing of humans. The study recognizes that urbanization often is inevitable and parts of the process are irreversible.
As the urban population increases, more buildings, services and infrastructure are needed. However, in most developing countries, vegetation spaces and structures are disappearing quickly at the expense of physical and population densification and poor planning. As a result, the vegetation is posed with the challenge of degradation and its eventual loss. The study seeks to analyses the urban expansion as compared to the rate of vegetation loss looking at Jimma town. Jimma Town in 1990 had a population of 2,486,155, persons as compared to 1,593,823 people in the year 2000. Thus, Jimma town recorded the highest percentage increase in population i.e. (38.1%) over the 30-years period (CSA, 2014). This implies that about 892,332 people have been added to the region Journal of Environment and Earth Science www.iiste.org ISSN 2224-3216 (Paper) ISSN 2225-0948 (Online) Vol.11, No.16, 2021 indicating a probable widening of the existing residential, educational and commercial areas. The expansion of these sectors cannot be done without the clearing of vegetated grounds leading to the loss of vegetation. In using GIS analysis and Remote sensing techniques, future predictions and trends of the urban land use and land cover and its subsequent development could be modelled and their results could help city planners and policy makers attain and sustain future urban development.

Objectives of the study
This project is aimed at Investigating the extent of vegetation loss through urban expansion in and around Jimma Town over a 30 period of 6 years interval by using Geographic Information System (GIS) and Remote Sensing (RS) techniques.

Specific objectives
ü To identify the various lands use and land cover (LULC) classes within and around Jimma town and their spatio-temporal distribution. ü To identify the association between urban expansion and vegetation loss.

Data Processing and Sources of data
The study employed both remote sensing techniques with GIS as the core components needed for such analysis.
In this project work, images of Jimma Town were downloaded via the Landsat remote sensor from the Earth explorer (www.earthexplorer.usgs.gov). The images downloaded were extracted and each layer stacked together using the digital image-processing software ENVI 5.2. The image classification method was adopted for the study. The process requires that images are put in classes and rectified. Supervised classification was employed in classifying the images from 1990,1996,2002,2008,2014 and 2020 into 5 hierarchical classes; Built up area ,Vegetation, Agricultural land , Grass land , and wetlands based on a modified classification scheme. The maximum likelihood classification algorithm was used in land use classification. Training samples for each image were plotted to ensure that different land use classes could be separated. This classification uses the training data by means of estimating means and variances of the classes, which are used to estimate probabilities and also consider the variability of brightness values in each class (Perumal & Bhaskaran, 2010).

RESULTS AND DISCUSSION
The study reveals that there is a remarkable urban expansion in the town especially in the periphery, over the past three decades (1990-2020).The built-up area was dominating the gains while grass land and agricultural land were dominating the losses and suggesting these changes may have been driven by a combination of biophysical, socioeconomic, and policy factors. Rapid urbanization and expansion of settlement may have further triggered the urban ecosystem, mostly gaining massive areas from agricultural land and forest over the past three decades . The result of the study indicated that there are spatial patterns of major LULC types in the study area ( Figure  3). A total of five LULC types were extracted in the study area over the three decade periods. Rapid urban expansion is clearly evident in Figure 3 and shrinking of agricultural land and grassland is also another fundamental LULC change in the study period 1990 and 2020. The overall accuracies for the Land cover map of the study area in the years 1990, 1996, 2002, 2008, 2014 and 2020 were 82.32%, 79.51%, 83.31%, 84.12%, 75.21% and 72.32% respectively. Producer's accuracy in 1990 land cover map ranged from the lowest value of 53% to the 86%, while in mapping LULC of year 2020 land cover map ranged from 72% to 97%. The least classification accuracy was observed for grass land and agricultural land (Table1, 2, 3, 4, 5 and 6) indicate details of accuracy results. 1990 1996 2002 2008 2014 2020 Fig. 3 The land use/land cover changes in Jimma city area during the study period from 1990 to 2020 Journal of Environment and Earth Science www.iiste.org ISSN 2224-3216 (Paper) ISSN 2225-0948 (Online) Vol.11, No.16, 2021       Land use/cover dynamics Land use/cover change analysis results and magnitude of land LULC change over the period 1990 and 2020 show that at the early time (1990), grass lands were the dominant land use land cover type, making up 39.03% of the study area. This was followed by agricultural land (33.20%), wetland (15.74%), built-up (13.16%), vegetation (8.84%), between years 1990and 2020, the LULC types have shown substantial positive and negative changes ( Table 7). Grass land, agricultural land and wetland were the top three LULC types, which have undergone large positive change of 116%, 69% and 66%, respectively. Whereas, vegetation have critically declined during these periods ( Table 7).The drastic consequences of ever-increasing demand for residential and institutional building construction spaces, associated poor urban planning(evidenced by widespread informal settlements) are mainly attributable to the observed changes. This study used past and recent satellite data to evaluate the land use change over the study period. For this purpose, five LULC were formed. These included; built up, vegetation, wetland, grass land and agricultural land. As indicated in Table 7, the highest LULC in1990 was grassland (39.03%). This was followed by agricultural land (33.20%) and wetland (15.74%). While built-up (13.16%) and vegetation (8.84%) is the least LULC classes in 1990. The analysis of LULCC revealed that the built-up area was 66.89 km 2 (13.16%) of the study area in 1990 was increased to about 385.83 km 2 in 2020. The agricultural land is decreased from 168.669 km2 in 1990 to 14.49 km2 in 2020 whereas, built up areas, the grassland and agricultural land was decreased by the year 2020. There has been a transformation from grassland land and agricultural lands to built-up area caused by rapid urbanization and slum proliferation (Mosammam et al. 2017;Esayas2013;Guan et al. 2011). In 2017, the LULC classification of Jimma city indicates that the highest proportion is vegetation land (30.65%), followed by barren land (21.23%) and open land (17.66%). The vegetation class shows a radical increasing in the study area because of the protection of the green area as well as the culture of the community in planting edible fruit trees and others. The built-up land cover class shows a remarkable increase between 2002 and 2020, which has been increased in size from 129.2985km2 in 2002 to 385.83 km2 in 2020, while the agricultural land was experiencing an extremely decreased from 98.973 km2 (19.48%) in 2002 to 14.49km 2 (2.852536 %) in 2020. The increment of built up area over the study period was associated with rapid population growth, migration